24I'll bet that if I was a Valve engineer working on tf2 and I had to stare at my screen at that code for any more than I already had for that day, I would probably say it would be time for a vacation.
The reason for Valve's delays is not because they're lazy, but because their employees are constantly encouraged to "push boundaries" and "invent new things", no matter how long it takes. If you're a coder at Valve and for four months straight, you report to your supervisor all the work you've been doing and it's nothing but TF2 fixes, they're going to question why you're even at the studio and probably fire you. But if you're a coder at Valve and for four months straight, you report this physics engine you're working on that has realistic player and environment reaction with simulated weight per prop, they're going to tell you it looks like it has great potential and keep it up.
They don't care if the physics coding takes eight years to perfect, they only care about being the studio that is constantly doing the next incredible that will sell big. It's what happened with Alyx; they had a team working on it for ages but when they finally finished, it caused a such a huge surge for the Index that they ran out of stock. THAT'S what Valve wants to see.
Any staff member who might be passionate about trying to fix TF2 is basically discouraged from doing so because there is no "progress" to be made from that.
https://i.imgur.com/6oPSXrq.png
Valve sees something like this and think, "We spent time on the Jungle Inferno update, didn't see a huge new peak in players, and instead, saw a big loss months later? What's the point of investing time to updating the game then?". Nevermind that the reason for the dip was because of big bugs (specifically for Pyro) that made the game unplayable. Valve just sees those numbers and thinks the investment of big updates is not worth it.